Endothelial cell dysfunction leading to decreased nitric oxide (NO) production is thought to be a key factor in the development of many types of cardiovascular disease. NO is produced from L-arginine in endothelial cells by the enzyme synthase (eNOS). Arginase, another enzyme found in endothelial cells, may compete with eNOS for arginine, which would be predicted to lead to decreased NO production. Arginase exists in two isoforms, arginase I and II, but it has not been definitively established which isoform may lower NO levels in endothelial cells and what the exact physiologic consequences would be. In order to examine this issue, transgenic mice over expressing either human arginase I (hArgI) or II (hArgII) gene under the control of endothelial-specific Tie2 promoter were produced. Both strains of transgenic mice had elevated arginase activity in endothelial cells, with no increased expression evident in macrophages. Over expression of hArg I or II did not lead to significant changes in plasma L-arginine levels or changes in vascular reactive oxygen species. Using aortic ring studies, Tie2hArgII, but not Tie2hArgI mice were found to have, however, diminished dilator response to acetylcholine. Furthermore, when Tie2hArgII mice were crossed with apoE knockout mice, they had a marked increase in aortic atherosclerosis on a normal chow diet compared to just apoE knockout mice. Collectively, these results suggest that therapeutic strategies involving the inhibition of hArgII may be useful for preventing endothelial dysfunction and atherosclerosis.